Deep Learning in Searching the Spectroscopic Redshift of Quasars

Abstract

Studying the cosmological sources at their cosmological rest-frames is crucial to track the cosmic history and properties of compact objects. In view of the increasing data volume of existing and upcoming telescopes/detectors, we here construct a 1--dimensional convolutional neural network (CNN) with a residual neural network (ResNet) structure to estimate the redshift of quasars in Sloan Digital Sky Survey IV (SDSS-IV) catalog from DR16 quasar-only (DR16Q) of eBOSS on a broad range of signal-to-noise ratios, named \code{FNet}. Owing to its $24$ convolutional layers and the ResNet structure with different kernel sizes of $500$, $200$ and $15$, FNet is able to discover the "\textit{local}" and "\textit{global}" patterns in the whole sample of spectra by a self-learning procedure. It reaches the accuracy of 97.0$\%$ for the velocity difference for redshift, $|\Delta \nu |<6000km/s$ and 98.0$\%$ for $|\Delta \nu |<12000km/s$. While \code{QuasarNET}, which is a standard CNN adopted in the SDSS routine and is constructed by 4 convolutional layers (no ResNet structure), with kernel sizes of $10$, to measure the redshift via identifying seven emission lines (\textit{local} patterns), fails in estimating redshift of $\sim 1.3\%$ of visually inspected quasars in DR16Q catalog, and it gives 97.8$\%$ for $|\Delta \nu |<6000km/s$ and 97.9$\%$ for $|\Delta \nu |<12000km/s$. Hence, FNet provides similar accuracy to \code{QuasarNET}, but it is applicable for a wider range of SDSS spectra, especially for those missing the clear emission lines exploited by \code{QuasarNET}. These properties of \code{FNet}, together with the fast predictive power of machine learning, allow \code{FNet} to be a more accurate alternative for the pipeline redshift estimator and can make it practical in the upcoming catalogs to reduce the number of spectra to visually inspect.

arXiv comment10 pages, 8 figures. The manuscript is accepted for publication in MNRAS

References in corpus (45)

• Machine learning and the physical sciences
• Empirical Evaluation of Rectified Activations in Convolutional Network
• The Baryon Oscillation Spectroscopic Survey of SDSS-III
• A luminous quasar at a redshift of z = 7.085
• An 800-million-solar-mass black hole in a significantly neutral Universe at redshift 7.5
• The Sloan Digital Sky Survey Quasar Catalog V. Seventh Data Release
• Massive Black Holes as Population III Remnants
• The SDSS-IV extended Baryon Oscillation Spectroscopic Survey: Overview and Early Data
• Spectral Classification and Redshift Measurement for the SDSS-III Baryon Oscillation Spectroscopic Survey
• The Assembly of the First Massive Black Holes
• BAT AGN Spectroscopic Survey - V. X-ray properties of the Swift/BAT 70-month AGN catalog
• The Sloan Digital Sky Survey Quasar Catalog: Sixteenth Data Release
• The Sloan Digital Sky Survey Quasar Catalog: Fourteenth Data Release
• A Survey of z>5.7 Quasars in the Sloan Digital Sky Survey IV: Discovery of Seven Additional Quasars
• A Luminous Quasar at Redshift 7.642
• Data Mining and Machine Learning in Astronomy
• The Fourth Catalog of Active Galactic Nuclei Detected by the Fermi Large Area Telescope
• Pōniuā'ena: A Luminous $z=7.5$ Quasar Hosting a 1.5 Billion Solar Mass Black Hole
• Eddington-limited accretion and the black hole mass function at redshift 6
• Star-galaxy Classification Using Deep Convolutional Neural Networks
• ANNz2 - photometric redshift and probability distribution function estimation using machine learning
• The Sloan Digital Sky Survey Quasar Catalog I. Early Data Release
• An Application of Deep Neural Networks in the Analysis of Stellar Spectra
• Automated Classification of Stellar Spectra. II: Two-Dimensional Classification with Neural Networks and Principal Components Analysis
• The Sloan Digital Sky Survey Stripe 82 Imaging Data: Depth-Optimized Co-adds Over 300 Deg^2 in Five Filters
• Estimation of stellar atmospheric parameters from SDSS/SEGUE spectra
• Automatic Redshift Determination by use of Principal Component Analysis --- I: Fundamentals
• Application of Convolutional Neural Networks for Stellar Spectral Classification
• High-redshift quasars and their host galaxies I: kinematical and dynamical properties and their tracers
• Deep learning Approach for Classifying, Detecting and Predicting Photometric Redshifts of Quasars in the Sloan Digital Sky Survey Stripe 82
• Probabilistic multi-catalogue positional cross-match
• Machine learning Applied to Star-Galaxy-QSO Classification and Stellar Effective Temperature Regression
• Machine Learning based photometric redshifts for the KiDS ESO DR2 galaxies
• Redshift Measurement and Spectral Classification for eBOSS Galaxies with the Redmonster Software
• QuasarNET: Human-level spectral classification and redshifting with Deep Neural Networks
• Optimal strategies for identifying quasars in DESI
• The newborn black hole in GRB 191014C proves that it is alive
• Parameterizing Stellar Spectra Using Deep Neural Networks
• On the Absence of High-Redshift AGNs: Little Growth in the Supermassive Black Hole Population at High Redshifts
• Spectroscopic QUasar Extractor and redshift (z) EstimatorSQUEzE I: Methodology
• Deep Learning for Multi-Messenger Astrophysics: A Gateway for Discovery in the Big Data Era
• Identification confusion and blending concealment in the SDSS-DR16 Quasar catalogues -- 40 new quasars and 82 false quasars identified
• Spectroscopic QUasar Extractor and redshift (z) EstimatorSQUEzE II: Universality of the results
• Tuned Inception V3 for Recognizing States of Cooking Ingredients
• Cosmological test using the high-redshift detection rate of FSRQs with the Square Kilometer Array

Cited by in corpus (7)

• Data mining techniques on astronomical spectra data. II : Classification Analysis
• Resolving the Vicinity of Supermassive Black Holes with Gravitational Microlensing
• Machine learning technique for morphological classification of galaxies from the SDSS. III. Image-based inference of detailed features
• Galaxy Spectra Networks (GaSNet). III. Generative pre-trained network for spectrum reconstruction, redshift estimate and anomaly detection
• FNet II: Spectral Classification of Quasars, Galaxies, Stars, and broad absorption line (BAL) Quasars
• Accurately Estimating Redshifts from CSST Slitless Spectroscopic Survey using Deep Learning
• Predicting large scale cosmological structure evolution with generative adversarial network-based autoencoders